Traffic Transformer: Transformer-based framework for temporal traffic accident prediction

<p dir="ltr">Reliable prediction of traffic accidents is crucial for the identification of potential hazards in advance, formulation of effective preventative measures, and reduction of accident incidence. Existing neural network-based models generally suffer from a limited field of...

Full description

Saved in:
Bibliographic Details
Main Author: Mansoor G. Al-Thani (19237204) (author)
Other Authors: Ziyu Sheng (19237207) (author), Yuting Cao (4231810) (author), Yin Yang (35103) (author)
Published: 2024
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1864513510219710464
author Mansoor G. Al-Thani (19237204)
author2 Ziyu Sheng (19237207)
Yuting Cao (4231810)
Yin Yang (35103)
author2_role author
author
author
author_facet Mansoor G. Al-Thani (19237204)
Ziyu Sheng (19237207)
Yuting Cao (4231810)
Yin Yang (35103)
author_role author
dc.creator.none.fl_str_mv Mansoor G. Al-Thani (19237204)
Ziyu Sheng (19237207)
Yuting Cao (4231810)
Yin Yang (35103)
dc.date.none.fl_str_mv 2024-04-01T03:00:00Z
dc.identifier.none.fl_str_mv 10.3934/math.2024617
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/Traffic_Transformer_Transformer-based_framework_for_temporal_traffic_accident_prediction/26389384
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Information and computing sciences
Artificial intelligence
Data management and data science
Machine learning
Mathematical sciences
Applied mathematics
traffic accident prediction
deep learning
transformer
attention mechanism
neural network
dc.title.none.fl_str_mv Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p dir="ltr">Reliable prediction of traffic accidents is crucial for the identification of potential hazards in advance, formulation of effective preventative measures, and reduction of accident incidence. Existing neural network-based models generally suffer from a limited field of perception and poor long-term dependency capturing abilities, which severely restrict their performance. To address the inherent shortcomings of current traffic prediction models, we propose the Traffic Transformer for multidimensional, multi-step traffic accident prediction. Initially, raw datasets chronicling sporadic traffic accidents are transformed into multivariate, regularly sampled sequences that are amenable to sequential modeling through a temporal discretization process. Subsequently, Traffic Transformer captures and learns the hidden relationships between any elements of the input sequence, constructing accurate prediction for multiple forthcoming intervals of traffic accidents. Our proposed Traffic Transformer employs the sophisticated multi-head attention mechanism in lieu of the widely used recurrent architecture. This significant shift enhances the model's ability to capture long-range dependencies within time series data. Moreover, it facilitates a more flexible and comprehensive learning of diverse hidden patterns within the sequences. It also offers the versatility of convenient extension and transference to other diverse time series forecasting tasks, demonstrating robust potential for further development in this field. Extensive comparative experiments conducted on a real-world dataset from Qatar demonstrate that our proposed Traffic Transformer model significantly outperforms existing mainstream time series forecasting models across all evaluation metrics and forecast horizons. Notably, its Mean Absolute Percentage Error reaches a minimal value of only 4.43%, which is substantially lower than the error rates observed in other models. This remarkable performance underscores the Traffic Transformer's state-of-the-art level of in predictive accuracy.</p><h2>Other Information</h2><p dir="ltr">Published in: AIMS Mathematics<br>License: <a href="https://creativecommons.org/licenses/by/4.0" target="_blank">https://creativecommons.org/licenses/by/4.0</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.3934/math.2024617" target="_blank">https://dx.doi.org/10.3934/math.2024617</a></p>
eu_rights_str_mv openAccess
id Manara2_5a458caaa428ed2ff49ce9179fe46492
identifier_str_mv 10.3934/math.2024617
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/26389384
publishDate 2024
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Traffic Transformer: Transformer-based framework for temporal traffic accident predictionMansoor G. Al-Thani (19237204)Ziyu Sheng (19237207)Yuting Cao (4231810)Yin Yang (35103)Information and computing sciencesArtificial intelligenceData management and data scienceMachine learningMathematical sciencesApplied mathematicstraffic accident predictiondeep learningtransformerattention mechanismneural network<p dir="ltr">Reliable prediction of traffic accidents is crucial for the identification of potential hazards in advance, formulation of effective preventative measures, and reduction of accident incidence. Existing neural network-based models generally suffer from a limited field of perception and poor long-term dependency capturing abilities, which severely restrict their performance. To address the inherent shortcomings of current traffic prediction models, we propose the Traffic Transformer for multidimensional, multi-step traffic accident prediction. Initially, raw datasets chronicling sporadic traffic accidents are transformed into multivariate, regularly sampled sequences that are amenable to sequential modeling through a temporal discretization process. Subsequently, Traffic Transformer captures and learns the hidden relationships between any elements of the input sequence, constructing accurate prediction for multiple forthcoming intervals of traffic accidents. Our proposed Traffic Transformer employs the sophisticated multi-head attention mechanism in lieu of the widely used recurrent architecture. This significant shift enhances the model's ability to capture long-range dependencies within time series data. Moreover, it facilitates a more flexible and comprehensive learning of diverse hidden patterns within the sequences. It also offers the versatility of convenient extension and transference to other diverse time series forecasting tasks, demonstrating robust potential for further development in this field. Extensive comparative experiments conducted on a real-world dataset from Qatar demonstrate that our proposed Traffic Transformer model significantly outperforms existing mainstream time series forecasting models across all evaluation metrics and forecast horizons. Notably, its Mean Absolute Percentage Error reaches a minimal value of only 4.43%, which is substantially lower than the error rates observed in other models. This remarkable performance underscores the Traffic Transformer's state-of-the-art level of in predictive accuracy.</p><h2>Other Information</h2><p dir="ltr">Published in: AIMS Mathematics<br>License: <a href="https://creativecommons.org/licenses/by/4.0" target="_blank">https://creativecommons.org/licenses/by/4.0</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.3934/math.2024617" target="_blank">https://dx.doi.org/10.3934/math.2024617</a></p>2024-04-01T03:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.3934/math.2024617https://figshare.com/articles/journal_contribution/Traffic_Transformer_Transformer-based_framework_for_temporal_traffic_accident_prediction/26389384CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/263893842024-04-01T03:00:00Z
spellingShingle Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
Mansoor G. Al-Thani (19237204)
Information and computing sciences
Artificial intelligence
Data management and data science
Machine learning
Mathematical sciences
Applied mathematics
traffic accident prediction
deep learning
transformer
attention mechanism
neural network
status_str publishedVersion
title Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
title_full Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
title_fullStr Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
title_full_unstemmed Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
title_short Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
title_sort Traffic Transformer: Transformer-based framework for temporal traffic accident prediction
topic Information and computing sciences
Artificial intelligence
Data management and data science
Machine learning
Mathematical sciences
Applied mathematics
traffic accident prediction
deep learning
transformer
attention mechanism
neural network